7/21/2014

1

AAPM Meeting 2014

3D Printed Phantoms for Small Field

Dosimetry Applications

Julian Perks, Ph.D. and Stanley Benedict, Ph.D.

U.C. Davis Radiation Oncology

Background

• End to end test of GK Perfexion

– Lack of test object / phantom for the Leksell G frame

and imaging fiducial systems (CT and MRI)

– RPC and commercial phantoms check dosimetry but

not accuracy of fiducials

– Elekta performs QA on G frame only annually

• Small animal irradiation

– Accuracy of dose delivery on linear accelerator

Abstract

• 3D scanning and printing technology is utilized to create phantom

models in order to assess the accuracy of ionizing radiation dosing

for two scenarios involving small field dosimetry.

• Firstly, an end to end test of the Gamma Knife Perfexion system is

performed.

• Secondly phantoms are designed to simulate a range of research

questions including irradiation of lung tumors and primary

subcutaneous or orthotopic tumors for immunotherapy

experimentation in mice. The mouse phantoms are used to measure

the accuracy of dose delivery and then refine it to within 1% of the

prescribed dose.

7/21/2014

2

3D printing resources

• Dedicated biomedical engineering laboratory

– UC Davis has BME department

• Nextengine 2020i laser scanner

– Digitizes 3D object, creating map

• Netfabb v. 5.0 software and Autodesk Inventor

Professional 2014 software

– 3D computer aided design (CAD)

• Objet 260 Eden printer – prints clear, opaque and rubber

3D printing resources cont.

• VeroClear photo-polymer

– 16µm layers, with each layer hardened by ultraviolet

light. Stated accuracy of the print is 20 – 85 µm for

features below 50mm and 200µm for the entire printed

object

• Kern Electronics Micro 24, 190W CO2 Laser Cutter

3D laser scanner

7/21/2014

3

3D Objet 260 printer

Kern Electronics laser cutter

Gamma Knife Perfexion QA

• End to end test for MRI based treatment

• To test MRI the phantom must have predefined

stereotactic coordinates

• Stereotactic frame must attach to phantom with sub-

millimeter precision

• MRI compatible markers at predefined Leksell

coordinates

7/21/2014

4

Printing G frame – proof of principle

• Confirm accuracy of laser scanning and printing

• Created duplicate G frame

– Laser scanning

– 3D printing

• Caliper measurements to compare printed to original

– Dimensions measured to within 0.01mm

• Confirmation by fit into GK Perfexion unit

Original G frame component

3D printed components

7/21/2014

5

Gamma Knife Phantom Process

• Built MR fiducial system in CAD software to be

mounted on G frame (virtual)

• Designed head shaped phantom

– Water filled

– Reduced weight by removing chin

– Two halves to allow printing shell

• Replaceable marker balls serve as stereotactic

coordinates

• Phantom includes detector chamber hole and film holder

Replaceable pin mounting sites (pads) and

screw in marker balls

Mounting plate

• Needed to fix spatial relationship between G frame and

phantom

– Fitting to human patient has spatial variability

– Mounting plate fixes variability

– Reproducible

– Built Leksell coordinate system in CAD software,

which means exact distances for marker ball to

mounting plate

7/21/2014

6

Printed Phantom on laser cut mounting plate

Phantom on plate ready to be fixed to G frame

7/21/2014

7

Phantom in frame lifted off mounting plate

ready for MR or treatment

X,Y,Z coordinate system matches Leksell planning system

7/21/2014

8

Leksell MRI fiducial box attached to printed G frame

7/21/2014

9

Details of image acquisation

• CT – GE lightspeed big bore

– 2mm slice thickness, helical SRS brain protocol

• MR – GE Signa 1.5T

– 2mm slice thickness, Gamma Knife T1 axial protocol

Screen capture CT contour on MR scan to assess systematic error

CT scan of printed phantom with dose distribution

Film

holder

7/21/2014

10

Gamma Knife Phantom Results – Leksell

coordinates in mm

MRI CT Difference

(MR - CT) Planned

100 marker x 100.1 x 100 0.1 x 100

y 97.7 y 98.4 -0.7 y 100

z 100.2 z 100.8 -0.6 z 100

140 marker x 66.2 x 66.5 -0.3 x 65

y 57.4 y 58.3 -0.9 y 60

z 60.8 z 60.8 0 z 60

160 marker x 125.3 x 125.6 -0.3 x 125

y 126.5 y 127.2 -0.7 y 130

z 41.2 z 41.5 -0.3 z 40

Gamma Knife Phantom Results -

Positioning

• Position of marker balls

• Less than 1mm difference between MR and CT for the

center of the marker

• Approximately 2mm shift of marker ball posteriorly

along y axis

• Further investigation underway on reproducibility –

suspect shift introduced tightening pins

Gamma Knife Phantom Results -

Dosimetry

• Absolute dosimetry with ion chamber

• Dose measurement

– Mean dose to chamber contour in planning system –

5.90Gy

– Measured dose, calibrated A1SL chamber, Max4000

electrometer – 5.94Gy

– 0.7% difference

• Further investigation will measure spatial dosimetry with

radiochromic film

7/21/2014

11

Small animal irradiation

• Radiation oncology department serves 8 investigators

with small animal irradiations

• Linear accelerator for human use

• Special calculations for small field applications

• Question of accuracy of dose prescription to small animal

Small Animal Printing Process

• 3D scan of toy mouse

• CAD model adjusted for each scenario

– Whole body irradiation

– Lung model

– Bilateral flank tumors

• Printed mice adjusted to accommodate A1SL or

MOSFET detectors

CAD model of mouse

7/21/2014

12

CAD model with lungs – bulk approach to density correction by removing 2/3 printed material

3D printed mouse with lungs and A1SL chamber

Printed mouse with bilateral tumors with MOSFETs

7/21/2014

13

Comparison 3D printed to real mouse – electron density

Scanned area Density g/cm3

Solid water 1.06

Bolus material (superflab) 1.00

Phantom mouse material 1.15

Phantom mouse (lung) 1.06

Mouse gut 1.12

Mouse lung 0.66

Mouse bone 1.21

Setup of Phantom

Irradiation

Energy Prescribed

dose /

monitor units

Measured

dose / Gy

Comments

Whole body, 1cm bolus

directly on mouse

6MV X-rays 2Gy / 191MU 2.001

Whole body in cage, 1cm

bolus material draped over

cage

6MV X-rays 2Gy / 191MU 1.968 1.7% lower due to loss

indirect bolus

Mouse gut, 1cm bolus, half

blocked field

6MV X-rays 2Gy / 203MU 2.003 Average measured dose

from three positions

Mouse lung, solid mouse,

1cm bolus, 3x3cm field size

9MeV electrons 2Gy / 227MU 1.958 2.1% lower than

prescribed

Mouse lung, solid mouse,

1cm bolus, 3x3cm field size

9MeV electrons 2Gy / 223MU 2.001 Monitor units adjusted

Mouse Dosimetry Results

Summary and Conclusion

• Demonstrate ability to create and model GK components with sub mm accuracy

• Printing options for wide range of detectors

• Mouse model allows validation of new research protocols